The Pricing Actuary’s Lens

Use a technology signal to ask better questions about claims, pricing and customer outcomes.

WORKED EXAMPLE · DRIVER ASSISTANCE

What would emergency braking change?

Advanced driver assistance systems (ADAS) can change both collision avoidance and repair costs.

Mechanism illustration

Technology signal

Autonomous emergency braking (AEB)

Detects an imminent collision and can brake automatically.

Available: The system is offered on the model.
Fitted: This vehicle has the relevant equipment.
Active: The system operates when the collision risk arises.

A capability rating alone does not establish fitment or use on the vehicles you insure.

Inspect the released Front AEB evidence

Frequency · claims or crashes: Reduction supported
Severity · average claim cost: Mixed; remaining claims may cost more
Recorded evidence strength: Strong
Direct UK evidence: Limited

Multiple real-world studies support lower crash or claim frequency. Transfer to UK portfolio pricing still requires care.

IIHS: front collision warning and automatic braking study (2018)

US police-reported rear-end crashes, comparing equipped and unequipped versions of the same models. Supports frequency direction; it does not measure UK repair costs.

Compare the other ADAS studies

02 Risk mechanism

Two paths to collision claims cost

Mechanism illustration · compare the same insured group and period.

Collision avoidance

Possible effect on frequencyFewer collisionsSome crashes may be prevented.

Sensors & calibration

Possible effect on severityMore costly repairsEach repair may be more complex.

03 · Economic outcome

The net cost still needs to be measured

Number of collision claims × average cost per claim

Total collision claims cost · same insured group and period.

Compare equipped and unequipped vehicles with similar vehicle and driver characteristics, driving conditions and insured exposure. Use developed claims from the same period.

How to compare the insurance outcome

Frequency means claims per unit of insured exposure, such as a vehicle-year. Severity means the average cost of those claims.

Count × average cost gives the total cost of the collision claims in that group and period. To compare groups with different amounts of insured driving, first express the claims count per vehicle-year or per insured mile.

Include repair costs and the mix of remaining claims. Public road-collision trends provide context; they do not measure the net cost to your insured portfolio.

Practical response

Ask the insurer’s next question

Test

On the vehicles we insure, do avoided collisions outweigh the additional cost of repairs?

Evidence to change the assessment: Reassess with verified fitment and exposure-adjusted UK claims experience.

The illustration establishes no numeric UK portfolio saving or pricing adjustment.

Evidence dates & releaseEvidence published 4 Oct 2026

Evidence period: Periods vary by sourceThe time measured. Publication and event dates appear with the individual sources.
Assessment changed: Change date unavailable in this releaseLast recorded material change to the insurance assessment.
Explainer changed: Change date unavailable in this releaseLast recorded change to its evidence or explanation. A routine source check does not reset it.
Source checks: 21 Sept 2026 – 26 Sept 20269 of 9 sources have a check date in this release. These dates describe inspection, not new evidence.

Approved evidence published 4 Oct 2026This topic, its downloads and linked Data Lab views use the same selected release.

Inspect the sources and their dates

FROM ANALYSIS TO A RESPONSE

State five conclusions

The ADAS assessment uses the same five dimensions as every topic.

DirectionWhich way could the risk move?: Some systems reduce collision frequency; sensor repairs can increase severity.
MagnitudeHow large could the effect be?: System-specific study effects are not a UK portfolio discount.
TimingWhen could it affect the insured portfolio?: New fitment precedes material whole-fleet penetration.
EvidenceWhat supports the assessment?: Strength varies by system, outcome, study design and UK transferability.
ResponseWhat should the insurer do next?: Test

THE QUESTIONS BEHIND THE ASSESSMENT

Apply the eight tests

Start with the loss mechanism, then check the market, capital and customer consequences.

Risk mechanicsWhat changes the losses we insure?4 tests

Frequency

How could this change the number of claims per unit of insured exposure?

ADAS applicationCollision-avoidance systems should put downward pressure on the crash types they can detect and mitigate. The insured effect will lag new-car fitment.

Severity

How could this change the average claim cost and the mix of smaller and larger claims?

ADAS applicationSensors, windscreens, calibration and electronic modules can raise the cost of otherwise repairable damage, while the mix of remaining collisions may also change.

Exposure basis

Are we measuring the right amount of risk: for example, insured miles or vehicle-years?

ADAS applicationVehicle-year remains observable, but model generation and verified capability may increasingly belong in the risk description.

Tail and accumulation risk

Could there be very large losses, or many losses from a shared cause?

ADAS applicationShared software, common suppliers and over-the-air updates create plausible correlated scenarios that historical vehicle-level experience may miss.

Market and capitalWhat changes pricing, selection and resilience?2 tests

Price response and selection

How does demand change with price, and how do customers joining or leaving change the risk mix?

ADAS applicationInsurers must recognise capability without over-crediting uncertain fitment or losing competitiveness through classifications customers cannot understand.

Capital vulnerability

Could losses use up capital before credible claims experience reveals an unsuitable assumption?

ADAS applicationHistoric frequency and severity relativities may become stale at different speeds across vehicle generations.

Customer outcomesWhat does this mean for the people we insure?2 tests

Customer expectation

What do customers expect to change, and will they recognise and pay for the value?

ADAS applicationDrivers may expect recognised safety technology to influence price, claims handling and repair quality—even when public fitment data remains incomplete.

Duty to act

Where Consumer Duty applies, could doing nothing allow foreseeable harm, poor value or another poor customer outcome?

ADAS applicationWhere material safety differences are known, ignoring them indefinitely in product design, pricing or communication deserves explicit governance.

Consumer Duty: The duty-to-act test is a decision prompt, not a conclusion that inaction is automatically a breach. The relevant standard is whether a firm acts to deliver good outcomes and avoids, or where necessary mitigates, foreseeable harm. FCA Handbook PRIN 2A.2

Use the Lens on another topic

Identify the change, follow the risk mechanism, then state the evidence needed for a response.

Return to the Lens overview